Nothing is more frustrating than building or upgrading a PC, pressing the power button, watching the fans spin up, and then seeing nothing on screen — or worse, getting a blank monitor with your GPU’s fans not moving at all. When a modular PSU cable is not powering your GPU, the symptom usually looks like this: the system powers on, case fans and motherboard LEDs light up, but the graphics card stays dark, the display shows “No Signal,” or the GPU’s status LEDs (such as ASUS Q-LED or MSI EZ Debug indicators) hang on a white VGA light. In some cases the card boots fine into light loads and then crashes or shuts down the moment a game starts, because only some of its power connectors are actually delivering current.
The good news is that this problem is rarely dangerous, and in most cases it is caused by a seating or configuration issue you can resolve in under fifteen minutes with nothing more than a flashlight and a screwdriver. The one exception worth treating with urgency: if you smell burning plastic, hear crackling, or see discoloration on any connector, shut the system down immediately and do not power it back on until the cable and connectors are inspected. Otherwise, work through the fixes below in order — each one is safe, reversible, and ordered from easiest to most involved.
As an Amazon Associate we earn from qualifying purchases at no extra cost to you.
Our Top Picks
Likely Causes
| Cause | How to confirm | Fix difficulty |
|---|---|---|
| PCIe cable not fully seated at the GPU or PSU end | Inspect both ends; look for a visible gap between the connector housing and the socket | Easy |
| Wrong cable type used (PCIe vs. CPU/EPS cable swapped) | Check the printed label molded into the connector; 8-pin PCIe and 8-pin EPS look similar but are keyed and wired differently | Easy |
| Using a pigtail/secondary connector instead of the primary one | Check which daisy-chain leg is plugged in; some cards only draw from the primary connector correctly | Easy |
| Daisy-chained single cable on a high-draw GPU | Card boots but crashes under load; GPU has two or three separate 8-pin inputs fed by one cable | Moderate |
| Damaged or third-party modular cable with incorrect pinout | Cable came from a different PSU brand or model; pins do not match your PSU’s layout | Moderate |
| PSU modular port mislabeled or peripheral port used by mistake | Check the PSU’s rear panel labeling; some units label ports as PCIe, CPU, SATA/HDD rather than by color | Moderate |
Fix 1: Reseat the Cable at Both Ends
What causes it
Modular connectors can look seated while the latch has not actually clicked into place. Shipping, case installation, or even the weight of thick cables pulling on the connector can back a plug out just enough to break contact on one or more of the twelve-volt pins.
How to check
Power off the system and flip the PSU switch to Off. Press the PSU’s power button, then hold your case’s power button for ten seconds to drain residual charge. Unlatch and pull the PCIe cable free at the graphics card end, then inspect the socket for bent pins or debris. Do the same at the PSU end.
What to do
Push the cable firmly into the GPU connector until you hear and feel the latch click. Repeat at the PSU end. Tug gently on the cable (no more than a light pull) to confirm it is locked. Then power on and check whether the card’s fans spin or the display signal returns.
How to undo it
Simply press the release latch and pull the connector out again. Reseating leaves no lasting change to the system.
Fix 2: Verify You Are Using the Correct Cable Type
What causes it
Eight-pin PCIe and eight-pin EPS (CPU) connectors are keyed nearly identically and fit into each other’s sockets with moderate force on some cables. Plugging an EPS cable into a GPU, or a PCIe cable into the CPU header, delivers power on the wrong pins and prevents the card from working — and can damage it.
How to check
Read the printing molded into the plastic of each cable end. PCIe cables are usually labeled “PCI-E” and EPS cables “CPU.” If the text is worn off, compare the wire colors and pin arrangement against your PSU manual’s cable chart. Native 12V-2×6/12VHPWR cables on modern units are labeled similarly.
What to do
Swap in the correct cable. PCIe power goes from a port labeled PCIe on the PSU to the card’s power inputs. CPU power goes from a port labeled CPU or EPS to the motherboard’s CPU_8PIN/EPS12V header. Never force an 8-pin plug that resists — the keying should allow insertion with light, even pressure.
How to undo it
Unplug and re-swap the cables back to their original positions if needed. No settings are changed.
Fix 3: Try a Different Modular Port
What causes it
On many modular PSUs, individual ports can fail, or a port labeled for peripherals may have been used by mistake. Some high-wattage units also split their PCIe ports across internal rails or sub-groups, and one group may be inactive if a related setting is off.
How to check
With power off, confirm the cable is in a port explicitly labeled “PCIe” or “VGA” on the PSU’s rear panel — not one labeled “SATA,” “IDE/Peripheral,” or “MOLEX.” If your PSU has multiple PCIe ports, note which one you used.
What to do
Move the PCIe cable to a different port labeled PCIe on the PSU and retest. If your unit has a hybrid or multi-rail switch, make sure it is set per the manual (usually “Multi” or “Single” rail as documented for GPU loads). If the second port works, the first port’s socket is likely damaged — leave it unused.
How to undo it
Move the cable back to its original port. No persistent settings are altered.
Fix 4: Stop Daisy-Chaining on High-Draw Cards
What causes it
A single PCIe cable with two 8-pin connectors daisy-chained off it is rated for roughly 300 watts through its main leg. Cards that pull 300 watts or more — many modern mid-range and all high-end GPUs — should get one dedicated cable per connector. When the pigtail leg is the only one delivering power, the card may fail to boot or throttle and crash under load.
How to check
Look at your cable: if both 8-pin plugs hang off one wire run from a single PSU port, it is a daisy-chain. Check your GPU’s board power rating on the box or the manufacturer’s spec page. If the card has two or three power inputs and draws above 300 W, run dedicated cables.
What to do
Connect one separate PCIe cable from its own PSU port to each power input on the graphics card. If your PSU does not include enough native PCIe cables, use its included adapters only as documented in its manual — do not buy unlabeled third-party cables.
How to undo it
Reconnect the single daisy-chain cable if you later install a lower-power card. There is nothing to revert in software.
Fix 5: Inspect and Reseat the 12V-2×6 / 12VHPWR Connector
What causes it
On newer GPUs using the 16-pin 12V-2×6 connector, partial insertion is the leading cause of no power or intermittent shutdowns. The connector needs roughly two-thirds of its length fully home to make contact on all sense pins; if the sense pins are open, the card will not draw power as a safety measure.
How to check
With the system off, look at the connector from the side — there should be no visible gap between the plug body and the socket shroud, and the latch should be fully engaged. Gently but firmly push the plug home until it stops with an audible click.
What to do
Unplug the 16-pin plug, inspect the pins inside for melting, discoloration, or debris, then reinsert it squarely and press firmly until it clicks. Avoid sharp bends within 35 millimeters of the connector housing. Power on and test. If the plug shows any melting, stop using that cable and contact the card or PSU manufacturer for a warranty replacement.
How to undo it
Squeeze the latch and pull the plug straight out. Nothing else changes.
Fix 6: Test the GPU in Another PCIe Slot
What causes it
Occasionally the problem is not the cable at all but a marginal PCIe slot latch, a slightly warped card, or debris in the slot causing poor auxiliary power or data contact. This is uncommon but quick to rule out.
How to check
Power off, remove the card (press the PCIe slot’s release latch, usually at the rear-right of the slot, before pulling), and inspect the slot and card’s gold fingers for debris or damage.
What to do
Reinstall the card in the same slot first, ensuring it clicks in fully. If the issue persists, move it to the second full-length slot (often labeled PCIE2 or PCIE5 on the motherboard). Enter the BIOS afterward — usually by pressing Delete or F2 during POST — and confirm the primary display setting under something like Settings > IO Ports > Initiate Graphic Adapter is set to Auto or PCIe, then save and exit.
How to undo it
Return the card to the top slot and re-enter the BIOS to restore the original primary display setting.
Fix 7: Test the Card With a Known-Good Cable or in Another System
What causes it
If every connection checks out and the card still gets no power, either the cable, the PSU’s PCIe rail, or the card itself is faulty. Isolating which one requires a known-good part.
How to check
Borrow a native PCIe cable from the same PSU brand and model family, or move the graphics card to a second, working PC. Alternatively, test the PSU alone: many units can be jump-started by bridging the green PS_ON wire to a black ground wire on the 24-pin connector with a paperclip (PSU disconnected from the motherboard, switch on), then measuring the PCIe cable’s 12 V pins with a multimeter — you should read between 11.4 and 12.6 volts.
What to do
If a known-good cable fixes the problem, replace the faulty cable through your PSU manufacturer’s warranty or parts store. If the card fails in another working PC, begin an RMA with the GPU manufacturer. If the PSU’s 12 V rail is out of spec on the PCIe outputs, RMA the PSU.
How to undo it
Reinstall the original hardware configuration once testing is complete.
If Nothing Works
If you have confirmed correct cable types, fully seated connectors, dedicated cables where required, in-spec 12 V readings, and the card still fails in a second system, the fault almost certainly lies with the graphics card or the power supply itself. Contact the GPU manufacturer’s support with your test results — most major brands ask for the PSU wattage, cable configuration, and behavior under load. For power supplies, most reputable brands carry a warranty of seven to twelve years and will ship replacement modular cables or a replacement unit. Do not attempt to re-pin or modify modular cables from a different PSU model: pinouts vary between brands and even between product generations, and doing so can destroy the card instantly when powered on. Until the faulty part is identified, avoid repeated power-on attempts, as intermittent connections under load generate heat at the connector.
FAQ
Can I use a modular cable from a different PSU brand?
No. Modular cable pinouts are not standardized between brands or even between a brand’s own product lines. A cable that fits may route twelve volts to the wrong pins. Only use cables supplied with your exact PSU model or replacements sold by that manufacturer for your model.
Why does my GPU work in the BIOS but crash in games?
This pattern strongly suggests partial power delivery — typically a daisy-chained cable on a high-draw card or a partially seated connector. Light desktop loads fit within what one cable leg can supply; game loads do not. Apply Fix 4 and reseat all connectors.
Does it matter which end of the cable I connect first?
Functionally, no — but connecting the PSU end first, then routing and seating the GPU end last, lets you make the final latch click cleanly without cable tension pulling the plug back out. Avoid tight bends near either connector.
How do I know if my PSU’s PCIe port is dead?
If the same cable works in a different PCIe port on the same PSU, the first port is suspect. Confirm with a multimeter: with the PSU jump-started, a healthy PCIe cable should read between 11.4 and 12.6 volts on its 12 V pins. If a known-good cable reads zero or far out of range from a specific port, stop using that port and contact the manufacturer for warranty service.
